Hospitalizations are common and costly, representing the largest share of health care spending in the US at over 30% of national expenditures. Understanding the emergency department's (ED) role in contributing to hospitalizations would substantially improve the health and quality of care for patients requiring emergency services. The ED is now the chief source for hospitalizations in the US, and emergency providers currently determine if a patient requires hospitalization 400,000 times a day in 5,000 EDs, resulting in 20 million annual admissions. Yet despite this crucial role, ED hospitalization practices have been largely overlooked as a circumstance where unwarranted variation may exist. Emerging evidence suggests admission decisions may be, in part, discretionary and influenced by non-clinical factors. Rigorous confirmation of these preliminary findings would have enormous implications for the quality and cost of care given that the downstream effects of hospitalization are considerable. Understanding sources of variation in ED hospitalization practices is the first step towards reducing unnecessary admissions while simultaneously reducing potentially harmful discharges from the ED. To date, the causes of variability in ED hospitalization practices across clinical conditions and its consequences for patient outcomes and health care spending has not been extensively examined. The research proposed in this application is designed to advance the science in understanding the costs and consequences of variation in ED hospitalization practices for the US public. With the help and guidance of his mentorship team, this proposal is also intended to support the continued career development of Dr. Keith Kocher, an extremely promising emergency physician and clinician- investigator, within the field of emergency care health services research. During the period of support, he will pursue didactic instruction in several disciplines, including doctoral-level courses in advanced statistical and econometric methods, health care organization, and health policy. He will also pursue a health policy fellowship designed for health services researchers to explore and strengthen the connections between public policy and their research as well as develop critical leadership and advocacy skills. The research plan has three related specific aims. Aim 1 will determine the patient clinical and hospital factors associated with variability in ED hospitalizations. Aim 2 will determine how patient non-clinical factors influence variability in ED hospitalizations. Aim 3 will evaluate the effect of ED hospitalization patterns on outcomes and costs. This work will provide the applicant critical skills to advance as an independent investigator within emergency medicine health services research and lay the foundation for projects aimed at improving emergency care delivery for all Americans.